The goal of this proposal is to translate genomic sequence data into high-quality biological knowledge on microbial signal transduction. Signal transduction pathways control important cellular activities ranging from virulence and antibiotic resistance in bacterial pathogens to intracellular communication and coordination of complex cellular functions in humans. Signal transduction is one of the most problematic areas for current genome annotation protocols because of the high sequence variability of input and output domains and mosaic architecture of signal transduction proteins. We will achieve the goal of this proposal via high- throughput genome processing, sophisticated computational protein sequence analysis, and collaborations with leading experimental scientists. The central questions in the biology of signal transduction are: (1) What proteins comprise signal transduction pathways and (2) How the proteins transduce signals. To address these questions, we will first develop a high-throughput computational approach to improve function prediction for signal transduction proteins in microbial genomes combined with experimental validation of selected targets by our collaborators (SPECIFIC AIM 1). Second, we will develop an innovative computational approach to predict contact sites in interacting proteins within the best-studied signal transduction pathway (SPECIFIC AIM 2). These predictions will also be validated by our collaborators. A Knowledge Environment developed under SPECIFIC AIM 3 will integrate results obtained under this proposal and provide free access to the data and computational resources. In the long term, this research will allow us to generalize the computational approaches to signal transduction developed in simple microbial systems and extend them to complex eukaryotic systems, including those in humans. This research will also have an immediate impact on understanding the biology of human pathogens and antimicrobial drug design, and contribute to improvement of automated annotation in primary sequence databases.